Earlier this week, North Dakota reached a grim landmark as the pandemic spreads like wildfire across the country — 1 in every 800 people in the state has died of COVID-19. More than 1 in 10 residents has had the virus.
But who are these people? Even as the US gets deeper into the third wave of the pandemic, the demographic breakdown of who has been infected still contains large gaps, including in the predominantly white, rural states where cases are now skyrocketing. Despite the urgent need to prevent further tragedy during the holidays, there’s little way for officials to target their response without a clear picture of what’s happening on the ground.
Earlier waves of the coronavirus were defined by their outsize impact on Black, Latino and Native American people, who were more likely to be essential workers, live in more crowded conditions, lack access to healthcare or insurance, and have higher rates of underlying health conditions. Yet there continued to be inadequate case data to determine the full effect of the pandemic as it moved into rural states with larger white populations, like North Dakota, South Dakota, Wyoming, Iowa, and Nebraska. Health experts say these inequities persist across geographies, and significant blindspots — with some states missing race data for up to one-third of cases — raise questions about whether officials and the public understand the full impact the pandemic is now having on groups that are at greater risk, and whether they are receiving adequate outreach and resources.
Sen. Elizabeth Warren told BuzzFeed News: “While the [Centers for Disease Control and Prevention] and its partner agencies have made progress in publishing COVID-19 data, much more work remains.” Warren and Rep. Ayanna Pressley today called for an investigation into the racial inequities in the government’s response to the pandemic. Warren has also proposed legislation and urged improved data reporting on COVID-19 and race and ethnicity, correctional facilities, and data exchange between tribal nations. “The federal response must include better data collection between the CDC, its partner agencies, state and local health departments, and Indian Health Service, and Congress must act to make public health and demographic data reporting mandatory,” Warren said.
Jennifer Nuzzo, a senior scholar at the Johns Hopkins Center for Health Security, said, “We know that disparities exist, but in order to understand them and to understand, more importantly, how to prevent them, we need to have better data.”
In New York City, the epicenter of the devastating first wave, health officials released race and ethnicity data about COVID-19 cases in April. “Once they reported race and ethnicity data, they [immediately] shifted testing resources to these communities that were being shown to be hardest hit, especially Black and Hispanic communities. And that's why data matters,” said Utibe Essien, assistant professor at the University of Pittsburgh School of Medicine. “It's to really help inform policy, and more importantly, implementation of public health guidance.”
Even the limited information from states where the coronavirus is now spreading quickest shows the percentage of cases in the white population is lower than their state's overall demographics. For example, of all positive cases in North Dakota, 53% of people identified as white, even though white people make up 87% of the population; of positive cases in South Dakota, 67% identified as white, while white people are 82% of the population; in Wyoming, 56% of cases and 84% of the population; in Iowa, 64% of cases and 91% of the population; and in Nebraska, 50% of cases and 89% of the population.
This, however, is only part of the story, and the finer details remain unknown.
In North Dakota, about 37% of cases were of unknown race “due to that field not being completed during testing and/or by the individuals refusing to report,” according to Grace Njau, director of special projects and health analytics at the North Dakota Department of Health. Ethnicity data, which merely indicates whether a person is “Hispanic” or “not Hispanic,” is severely incomplete, with that information only available for 11% of active cases in the state, according to data Njau shared. “These data, and race data, are self-reported during testing,” she said.
It is not clear whether any racial group is overrepresented in this “unknown” category, and more complete data might show even greater disparities. “I don't have any reason to think that if we could identify all those unknowns that our understanding of disparities would totally change,” said Nuzzo. “My hunch is, based on historic access and equity issues, that it would probably just further slide the disparity a bit more to the extreme.”
Missing demographic data has been a problem since the spring, including in New York state (which continues to not report the racial breakdown of cases). But the country has been dealing with the pandemic for the greater part of the year now. In the national tally compiled by the CDC, race and ethnicity was still missing in about half of cases. “If at this point, we still don't feel confident that we have the full picture, it's really quite alarming,” said Essien, who coauthored a report about this issue earlier this year. “A lot of states still have dangerously high levels of missing information, to the point where it doesn't even feel comfortable that we can do anything sufficient with the data.”
As the virus moves into areas with far smaller populations of color that are more likely to be uninsured, unemployed, and live in poverty but have less visibility due to their size, it is important that they are not overlooked.
Essien said even in states with more racially homogeneous populations, there remains a strong need to track what is happening in minority populations — particularly when they are very small. He pointed to the example of Maine, where only about 1.7% of the population is Black. When the state released the racial breakdown of COVID-19 cases in the spring, it revealed Black residents made up 3.7% of cases, a share that has since grown to 11% of cases.
Because the unknown population remains so large, “it may be hard to look at what's going on in our minority communities in North Dakota,” said Brenton Nesemeier, a regional epidemiologist in Fargo for the North Dakota Department of Health. With more comprehensive data collection, health workers can offer more free testing events and do more precise outreach to community leaders, churches, and local business leaders, he said. “I think any amount of data is more helpful.”
Nesemeier said when cases began surging in North Dakota this fall, overwhelmed caseworkers condensed their interviews, collecting fewer close contacts and fewer details about people’s work, for instance. Contact tracers unable to keep up with the surge essentially gave up, making calls as quickly as possible and urging those who test positive to notify people they’re been around rather than doing that work themselves, Kailee Leingang, a nursing student and contact tracer in North Dakota, wrote in the Washington Post. “My understanding is, perhaps those questions [about race] just get overlooked sometimes,” Nesemeier said. “We need to go back and retrain to get those questions asked and ensure more accurate reporting of that information.” He described the size of the state’s “unknown” population as “something that we do need to improve on.”
This is no small challenge. Nesemeier cautioned that releasing demographic data is not without risks, especially in areas where the minority population is so small that they may feel singled out in their communities. “With North Dakota being a majority white, Caucasian state, sometimes when that higher-level race information is released, there is a stigma, and we want people of other races and minorities and other ethnicities to go in and get tested and feel like that it's OK to get tested. We don't want any stereotypes or any stigma coming down on them.”
Anecdotally, Essien said, Latino immigrants might feel uncomfortable documenting their race due to fear that it will limit their access to care. “And we have seen time and time again, Black Americans, when you don't list your race on your housing information or your loan information, you're more likely to get that loan or get that grant,” he added. “Perhaps individuals are making the same calculus with their health and saying, ‘Well, I'm not going to list my information here; maybe I'll be more likely to get that COVID test or more likely to get that vaccine.’”
While demographic data about people who have died from COVID-19 tends to be more complete, understanding case data earlier on can inform measures for prevention and treatment. While racial data about people who have died from COVID-19 was not available in North Dakota at the time of publication, Leingang told BuzzFeed News she has observed a clearer disparity in outcomes, especially among Native American patients who test positive for COVID-19.
In Wyoming, Clay Van Houten, infectious disease epidemiology unit manager at the state’s health department, said while the cases with missing race and ethnicity data do not appear to be coming from one county or provider or hospital group, “Our death data is more complete, and continues to show that a disproportionate number of American Indians have been affected by the pandemic in Wyoming.” Native Americans are just below 3% of Wyoming’s population but more than 11% of COVID-19 related deaths.
The health departments of South Dakota, Iowa, and Nebraska did not respond to inquiries from BuzzFeed News.
In addition to gaps in reporting racial demographics at the state-level, Keith Mueller, director of the Rural Policy Research Institute at the University of Iowa, said there is a lack of information about who is getting sick in rural areas specifically, making it unclear how health outreach can be more effective as the virus spreads outside of urban hubs. “We have not been able to secure a national data set that provides information on individuals testing positive by county,” he said.
Broadly looking at the worst-hit rural counties since February, Mueller said, they tend to have a larger nonwhite population, a larger Latino population, more people living in group quarters, and a larger uninsured population on average than rural counties with the lowest rates of COVID-19. “It's clearly some of the same population groups that we've heard about being disproportionately affected in urban areas are at risk in rural [areas] ... We can't get all the way down to the individual level, but we do know the counties that had the highest case rate,” Mueller said. At this late point, the absence of more granular information is “making analysis of rural populations impossible.”
John Templon contributed to this story.